This dataset contains vascular plant species cover and diversity indices (richness, evenness, and beta diversity) of understory communities and site-level environmental variables from 33 mature aspen (Populus tremuloides) stands across the Canadian boreal forest. The objective of the study was to evaluate the relative impacts of local and regional environmental conditions on vascular plant community composition and diversity in the boreal forest. We measured community composition and vegetation diversity indices of the vascular understory in 400 m² plots and measured environmental variables at the local (drainage, nutrient and light availability, site heterogeneity) and regional (annual temperature and precipitation, fire weather index, landscape heterogeneity) scales. 

Sampling took place in 33 trembling aspen (Populus tremuloides) stands along a transcontinental gradient in precipitation and fire regimes in boreal Canada. We measured community composition and vegetation diversity indices of the vascular understory in 400 m² plots and measured environmental variables at the local (drainage, nutrient and light availability, site heterogeneity) and regional (annual temperature and precipitation, fire weather index, landscape heterogeneity) scales. Specific details about data collection of individual variables are presented in the Usage Notes section. 

Associated paper: Crispo, Jean, Fenton, Leduc, and Bergeron. Factors explaining the composition and diversity of vascular plant understories along a transcontinental gradient in the Canadian boreal forest. Journal of Vegetation Science. Manuscript ID JVS-RA-05658. Under review

Authors names: Maude Crispo¹, Mélanie Jean^2,3 (ORCID: 0000-0002-7393-5566), Nicole J. Fenton² (ORCID: 0000-0002-3782-2361), Alain Leduc¹, Yves Bergeron^1,2 (ORCID: 0000-0003-3707-3687)

Author's institutional affiliations:

1 Département des sciences biologiques, Université du Québec à Montréal

2 Institut de recherche sur les forêts, Université du Québec en Abitibi-Témiscamingue

3 Département de biologie, Université de Moncton

Corresponding author: Mélanie Jean (melanie.jean@umoncton.ca)

Usage information: 

Species_cover sheet

Vascular plant species cover data at each site. One 400 m2 circular plot was randomly positioned in each of the 33 selected stands. In each circular plot, 20 quadrats (1 m2) were arranged in a cross shape pointing at the four cardinal directions with five quadrats on each branch. A survey of the vascular understory vegetation was carried out inside each quadrat by identifying all herbaceous and woody species and defining their percent cover based on vertical projections. Percentages were visually assessed by two independent observers and rare species (<1% cover) were assigned a cover of 0.5%. For each site, plant covers were summed up by species using the 20 quadrats (covers ranged from 0-2000). 

Site: Site code that contains the Canadian province it was in (AB: Alberta, BC: British Columbia, MB: Manitoba, ON: Ontario, QC: Quebec, SK: Saskatchewan) and an alphanueric code. 

Columns: Columns are alpharnumeric codes attributed to the species found in the surveys. Species names associated with the codes are in the Species_codes sheet. Unknown species that could not be identified have codes that start by "INC".

Species_codes sheet

CODE: Code used to identify the species in the Species_cover sheet. 

Genus : Genus of the species. 

Species: Species when known, otherwise the code "sp" was used to indicated that identification was done at the genus level only. 

Subspecies: When possible. 

Variety: When possible. 

Diversity_indices sheet

Site: Site code that contains the Canadian province it was in (AB: Alberta, BC: British Columbia, MB: Manitoba, ON: Ontario, QC: Quebec, SK: Saskatchewan) and an alphanueric code. 

Richness: Species richness encountered at a given site was calculated by counting the species inside and outside the quadrats within the 400 m2 plot. 

BetaDiversity :The total non-directional beta diversity within each site, or intrasite species heterogeneity, was determined by the total variance observed in species abundance data across the 20 quadrats at a site following Legendre and De Cáceres (2013) with the beta.div function in the adespatial package (Dray et al., 2017). 

Evar: Evenness was measured using the Evar index (Smith and Wilson, 1996), which is independent from richness and is not sensitive to rare species. 

Environmental_variables sheet

Sites: Site code that contains the Canadian province it was in (AB: Alberta, BC: British Columbia, MB: Manitoba, ON: Ontario, QC: Quebec, SK: Saskatchewan) and an alphanueric code. 

LAI: We calculated the leaf area index (LAI; inversely proportional to light availability) by processing hemispherical photographs taken at a height of 1 m in the centre of each plot and 7.5 m from the centre at the four cardinal points with Gap Light Analyzer 2.0 (Frazer et al., 1999). Photographs were taken after moving shrubs (˂4 m) out of the way. 

HET: A local heterogeneity index was calculated to represent the diversity of ecological niches at each site. This index characterized, in a single relative numerical value, dead wood or rock presence on the ground, microtopography, and presence of windfalls and windbreaks creating canopy gaps and altering the physical structure of the site (Appendix S1). Individually, these components had highly skewed distributions and seemed insufficient to fully represent ecological niche diversity, while the synthetic local heterogeneity index variable had a normal distribution. See Maude Crispo's Masters thesis (Université du Québec à Montréal) for more details on this variable. 

DRAIN: At each site, a drainage class ranging between 0 (excessive drainage) and 6 (very poor drainage) was established based on soil characteristics (Saucier et al., 1994). 

lnCEC: Cation exchange capacity (CEC) was determined by the Mehlich-3 method on mineral soil samples taken 15 cm below the organic layer, at the four cardinal points of the plots. This variable was transformed with the natural log to improve linearity in our models (CEC; see below). 

sqFWI: Fire susceptibility was quantified by the forest fire weather index (FWI), calculated from six components taking into account effects of fuel water content, wind, and weather conditions on fire behaviour, provides an estimate of forest fire occurrence risk (Wotton, 2009). FWI was square root transformed to respect assumptions of normality and homoscedasticity. Fire susceptibility was averaged over 30 years by the BioSIM 9 software (Régnière and Saint-Amant, 2008) by interpolating data from the closest weather stations to each site (1981-2010; Environment Canada, 2013). 

DD: Annual sum of degree days (> 0º C) at the site averaged over 30 years by the BioSIM 9 software (Régnière and Saint-Amant, 2008) by interpolating data from the closest weather stations to each site (1981-2010; Environment Canada, 2013).

LH: Landscape heterogeneity was calculated as the proportion of forest cover within a 1 km radius around the site (see Crispo [2016] for calculation and rationale details), thus accounting for the presence of species from open or disturbed environments, such as agricultural land, roads, or clearcuts.

RAIN: Annual precipitation as rain (mm) at the site,  averaged over 30 years by the BioSIM 9 software (Régnière and Saint-Amant, 2008) by interpolating data from the closest weather stations to each site (1981-2010; Environment Canada, 2013). 

Latitude: Site latitude in degree decimal. 

Longitude: Site latitude in degree decimal.